14
High-resolution simulations of air-sea interaction in the Mediterranean Sea R. Justin Small, Travis Smith, Tim Campbell, Rick Allard, Jim Dykes Naval Research Laboratory, Stennis Space Center Joao Teixiera, Jet Propulsion Laboratory, NASA Acknowledging Sue Chen, Jim Doyle, NRL Monterey AMS 16 th Conference on Air-Sea Interaction, Jan. 11-15 2008.

High-resolution simulations of air-sea interaction in the Mediterranean Sea R. Justin Small, Travis Smith, Tim Campbell, Rick Allard, Jim Dykes Naval Research

  • View
    216

  • Download
    0

Embed Size (px)

Citation preview

High-resolution simulations of air-sea interaction in the Mediterranean Sea

R. Justin Small, Travis Smith, Tim Campbell, Rick Allard, Jim DykesNaval Research Laboratory, Stennis Space Center

Joao Teixiera, Jet Propulsion Laboratory, NASA

Acknowledging Sue Chen, Jim Doyle, NRL Monterey

Julie Pullen NRL Monterey/Stevens Lab NY

AMS 16th Conference on Air-Sea Interaction, Jan. 11-15 2008.

Introduction

• Advantage of a mesoscale or regional coupled model for forecasting– Regional-Large scale influence constrained by boundary conditions– Mesoscale-Focus on smaller scales of interest (high resolution)– Coupled-Self-consistent local ocean, atmosphere and wave forecasts– Coupled Improve fluxes etc. by allowing components to interact

• Evolution of coupled boundary layers important for Naval operations– e.g. acoustic detection (Sonar): good fluxes and stresses needed for proper

forecasting of mixed layer depth, thermocline gradient, waves… – e.g. surface drift: surface currents – consistent wind, waves, currents required.

Aims• To validate the coupled model against experimental data.

• Adriatic Circulation Experiment February 2003• Ligurian Sea Air-Sea Interaction Experiment June/July 2007

• To investigate fine-scale wind events due to steep land topography and their effect on the ocean• Sensitivity to resolution• Sensitivity to coupling

• Starting with …(preliminary results)• Flux comparisons

Numerical Model• COAMPS® (Coupled Ocean/Atmosphere Mesoscale

Prediction System)– Non-hydrostatic, compressible, sigma coordinate– Explicit microphysics for ~10km or less grid space– Kain-Fritsch (1990, 1993) for >10km– Modified Louis (1979) surface layer

• NCOM (Navy Coastal Ocean Model)– Hydrostatic, incompressible, Boussinesq approximations– free-surface, mixed sigma(upper) and z-level(lower)– Mellor-Yamada vertical mixing

Coamps5 Regions -MediterraneanNorthern Adriatic SeaEvents of interest: winter-time Bora

Ligurian Sea

Events of interest: Alpine Lee cyclogenesis and strong winds in summer

Experimental design

• Atmospheric nests– 3 nests, 27 km, 9km, 3km grid spacing– 30 levels

• Ocean nests– 2 nests, 6km, 2km– 30 levels

• Initial and Boundary conditions– NOGAPS (2003, 1°, 2007 0.5°) 6 hourly– Global NCOM (1/8°) – Initialise 6 days before

Northern Adriatic BoraSurface fluxes and stress from observations and coamps: from Pullen et al (JGR 2007)

Data: Extensive previous studies include Adriatic Circulation Experiment (2002-2003). Stress and flux data for 1 month at gas platforms and land stations (Jan/Feb 2003 Dorman et al 2006), ship data. Previous Modeling: includes a file-based coupled coamps experiment (Julie Pullen, Jim Doyle, Tracey Haack). Aim: to reproduce and possibly extend their results with the ESMF-based coamps5. (e.g. utilising different coupling options, timesteps etc.).

Topography and Bora schematic

Bora Event Validation-wind stressSite Uncoupled

(New)Uncoupled Pullen et al 2007

CoupledPullen et al 2007

Venice 0.59 0.68 0.64

Azalea 0.63 0.52 0.46

Ancona 0.38 0.56 0.59

Veli Rat 0.77 0.71 0.69

Wind StressCorrelation coefficient based on 30 days of hourly data from station and from model

VENICE

ANCONA

AZALEAVELI RAT

Bora Event Validation-net heat fluxSite Uncoupled

(New)Uncoupled Pullen et al 2007

CoupledPullen et al 2007

Venice 0.79 0.55 0.57

Azalea 0.84 0.75 0.77

Ancona 0.61 0.69 0.7

Veli Rat 0.92 0.72 0.73

Heat FluxCorrelation coefficient based on 30 days of hourly data from station and from model

Net heat flux=long wave net (up)-incoming short wave+(sensible heat +latent heat out of ocean)

LASIE07 (PI: Joao Teixiera)Data: LASIE07 (June/July 2007, PI: Joao Teixiera). Over 300 ocean profiles,

100 radiosondes, 2 surface met moorings, wave buoys, drifters, ADCP, thermistor moorings.

Aim: to better understand coupled boundary layer processes.Example process: strong summer ‘Mistral’ winds cool SST: how deep does this

affect extend and is there a feedback to atmosphere? Impact on clouds, heat fluxes and stress. Change of wave field.

La SpeziaNATO Undersea Research Center (NURC)

Ligurian Sea ValidationSite 10m wind

speedSolar radiative flux

Long wave net flux

ODAS 0.57 0.96 0.8

METEO 0.46

Wind speed and radiative fluxesCorrelation coefficient based on 30 days of hourly data from station and from model

LASIE07: Synoptic event 25/26 JuneWind speeds and vectors at 10m, 26th June, 1700UTC. Left: COAMPS simulation, Bottom: QuikSCAT.

Ocean response (surface temperature)

COOLING

• SUMMARY– High resolution stand alone model runs performed for Adriatic and

Ligurian Sea– Fluxes from atmosphere-only run correlate with observations typically >

0.5.

• WAY AHEAD– Preliminary coupled runs being performed– Validation of model against in-situ profiles in the atmosphere and ocean– Investigation of coupled effects during strong wind events– Validation in other regions – against datasets such as KESS, CLIMODE,

CBLAST